During the previous funding period we began studying learning in the caudate and the globus pallidus internus. We more clearly elucidated that the caudate is involved in executive aspects of learning, rapdily making new associations between the visaul cues, specific motor movements, and reward. We discovered that intermittent electrical micro-stimulation of the caudate leads to enhanced learning beyond normal rates . We found that the globus pallidus internus plays a key role in learning by initially faciliatating exploration, while later maximizing consolidation of learned associations. These results were published in a series of articles in Nature Neuroscience and the Journal of Neuroscience. We also made a number of novel preliminary findings that form the basis of the current proposal. We found that, the nucleus accumbens has a very different pattern of neuronal activity than the caudate suggesting a role in motivation. We also found that intermittent electrical stimulation in the ventral striatum or nucleus accumbens seems to ehance motivation. We can now posit a two stream model of the anterior striatum, where the dorsal striatum is involved in the executive aspects of learning, while the ventral striatum plays a role in the motivation to learn. The goal of the current proposal is to explicitly address two hypotheses. The first hypothesis is that there are two streams of processing in the anterior striatum, one critical to learning and the other for motivation. In this regard, our goal is to demonstrate that he dorsal module, Caudate (Cd) and dorsal lateral prefrontal cortex (DLPFC), is involved in the executive aspects of learning, that is the rapid association of visual cues and motor responses, while the ventral module, nucleus accumbens core (NAcc) and orbital-frontal cortex (OFC), plays a complementary role, by providing the motivation to perform learning behavior. The second hypothesis is that the dorsal and ventral modules also play important and complementary roles in working-memory, which is an essential component of associative learning. In regard to working-memory, our hypothesis is that the Cd is critical in gating relevant visual images, while the NAcc is critical in linking reward value with remembered images. We will employ a multi-modal approach using neuron physiology, connectivity analysis, and microstimulation. These results will be extremely important for optimizing and generating treatments for a broad range of neurological and psychiatric disorders.